The FDA and Magical Thinking

Vox had a piece yesterday on the Cruz-Lee proposal to make it easier for U.S. patients to access drugs and devices already approved in other developed countries. The Vox piece had some howlers. Most notably this:

“There’s no evidence the FDA blocks innovation or makes innovation harder or makes it more costly,” said Kesselheim.

Frankly, that would be laughable were it not coming from a professor of medicine at Harvard Medical School. It costs well over a billion dollars to get the average new drug approved and much of that cost comes from FDA required clinical trials. Longer and larger clinical trials mean that the drugs that are eventually approved are safer. But longer trials also mean that good drugs are delayed. And the more expensive it is to produce new drugs the fewer new drugs will be produced. In short, longer and larger trials mean drug delay and drug loss.

We live in a world of tradeoffs. Let’s debate the tradeoffs. But let’s not engage in magical thinking where there are no tradeoffs and “no evidence” that the FDA makes drug development more costly.

A more subtle error was committed by the author who writes:

But it’s not clear that this legislation can solve the biggest problem here — the lack of promising treatments in the pipeline. In other words, a faster approval process can’t fix a dearth of innovation from labs themselves.

Many factors go into drug development that are outside the FDA’s purview. Nevertheless, faster drug approval can and does increase innovation. Approving drugs more quickly is equivalent to a decrease in the costs of research and development. Time is money. Reducing the cost of development increases the incentive to develop new drugs.

The Prescription Drug User Fee Act, for example, reduced drug approval times by about 10 months. Philipson et al. calculate that:

…the more rapid access of drugs on the market enabled by PDUFA saved the equivalent of 140,000 to 310,000 life years.

(PDUFA does not appear to have materially affected safety but Philipson et al. calculate that even under a worst case scenario the benefits of PDUFDA far exceeded the costs).

Moreover, Vernon et al. find that the reduction in approval time from PDUFA increased new drug development:

Controlling for other factors such as pharmaceutical profitability and cash flows, we estimate that a 10% decrease (increase) in FDA approval times leads to an increase (decrease) in R&D spending from between 1.4% and 2.0%. Combining this estimate with recent research on the link between PDUFA and FDA approval times…we calculate PDUFA may have incentivized an additional $10.8 billion to $15.4 billion in pharmaceutical R&D. Recent economic research has shown that the social rate of return on pharmaceutical R&D is very high; therefore, the social benefits of PDUFA (over and above the benefits of more rapid consumer access) are likely to be substantial.

Finally, return to the issue of reciprocity. Many of the critics of reciprocity respond with simple appeals to nationalism. We are the best! Rah, rah, rah! But if the critics were German or French they would argue that the EMA is superior to the FDA. Indeed, when I raise the issue of reciprocity with Europeans they respond in exactly the same way as Americans. How could anyone suggest that the EMA automatically approve drugs approved by the FDA! The horror.

The argument for reciprocity, however, isn’t that the FDA is uniquely bad or always worse than the EMA or vice-versa. The argument is that it’s wasteful to duplicate the lengthy approval process and that both agencies sometimes make mistakes. As a result, it’s simple common sense to let Americans avail themselves of drugs and devices approved in other developed countries.


The fact that Priority Review Vouchers can be sold for hundreds of millions of dollars each is prima facie evidence that the FDA retards innovation. Those vouchers only promise moving up in the FDA queue, and are rewarded for successfully researching a high priority tropical disease or other rare disease lacking treatment. Just for jumping higher in the queue it's worth that much! Good on the FDA and Congress for making them tradeable, though of course some usual suspects are trying to game them, getting approval for drugs approved elsewhere-- a problem to be solved by reciprocity rather than awarding vouchers. Vouchers are too high a reward for shepherding an existing drug as opposed to researching a new one.

Its only prima facie if you assume that "profitable" equates with "innovative", which it surely does in one sense but not necessarily the "saving lives" sense.

Drug companies spend a whole lot of money researching lifestyle drugs and demonstrating marginal improvements on existing drugs in order to renew their patents. These profit centers need approval just like other drugs, and Pfizer wanting their new boner pills approved more quickly doesn't mean the FDA is preventing cancer research.

Why do lifestyle drugs always get denigrated by those trying to retard innovation? The only drugs we should care about are ones that extend life, not those that improve it? Viagra is probably one of the most important drugs in the history of drug development. It basically made life worth living again for millions of people. Maybe you are some weird asexual person, but the vast majority of us place a lot of value on being able to have sex.

Who's trying to retard innovation?

The reason lifestyle drugs get denigrated is because they're 1) often the innovation that actually results form laxer standards and 2) never the type of innovation that's brought up by people pushing for reform. Making a marginally different NSAID could be a very worthwhile and useful endevour, but that's the context we should be thinking rather than curing cancers left and right.

Amen! I bet if there were a "lifestyle" drug that made you a better game player, our friend "mravery" would get it so he could be more popular with the people he meets on-line while playing games in his Mom's basement.

The generic drug that got its price raised by 1000% could be imported from China for pennies, while there was no secondary American producer.

I'm sure it would cost a bit more to run in America with American quality-control, so say nickels instead of pennies. So why was no one else in America making it?

I can think of a few reasons, but they all feel kind of flimsy compared to "not worth the regulatory headache to bring a new drug to market."

I agree with this, but I think you too deny a tradeoff at the end. There really is a risk of regulatory arbitrage and possible race to the bottom. I think that risk is over-stated, and the benefits greatly outweigh it. But reciprocity is not 'simple common sense'.

There's also an argument for legitimate national differences. Some polities might be more risk averse in certain respects (e.g. European mistrust of GMOs). So national differences might not be a result only of competence. I'm not making that argument myself - I'd be happy to take my chances with drugs approved by any major regulator. But reasonable people may disagree, as they say.

I mean, I would trust the UK, Sweeden, or France with whether a drug is okay or not, but is there really any evidence that Greece, Italy, Spain, etc. have the proper capacity and capability to deal with what could easily become a deluge of approval requests? Why do folks who hate our public bureaucracy think that Greece's will do a better job?

I think reciprocity is a good idea, but it should be implemented carefully. Temporary approval (3-5 years) with the requirement of subsequent FDA trials seems the logical approach to me. And maybe you trim the list from "everyone in the EU plus others" to something smaller.

"Why do folks who hate our public bureaucracy think that Greece’s will do a better job?"

Sigh. Greece doesn't have it's own independent drug approval agency. The European Medicines Agency handles drug approvals for the EU. And even if Greece *did* have its own independent agency, the US could simply exclude it from the reciprocity list. But small countries don't have their own independent versions of the FDA because pharma companies couldn't justify re-doing expensive clinical trials for small markets.

Most major medication approval is centralized in the EU:

For medication that is not subject to centralized approval, there is local approval with reciprocity, overseen by an EU-wide organization.

I think this is much less of a risk than you contend.

That is a lot less risky. It seemed ridiculous that such relatively small countries could maintain a bureaucracy large enough to deal with the scope that our FDA has to contend with. I still kinda wonder how Canada, Israel, and Australia do it.

This reinforces my opinion that reciprocity is a good idea, though I still think it should be temporary and require subsequent approval by our agency. It'd be nice if all of these countries could get together and have some kinda of "common application" so a manufacturer could just do them all at once....

What is the mechanism whereby basic reciprocity with certain countries triggers a "race to the bottom?" How does it benefit Country A to start watering down its pharmaceutical approval process just because Country B says it will accept Country's A's determinations? Maybe your concerns are valid, but I don't see why, and this "race to the bottom" language is, in my experience, frequently invoked by people of questionable economic literacy to defend minimum wage laws, which makes me quite skeptical of people invoking it other contexts.

Companies spend a lot of money on clinical trials.

That doesn't answer my question.

I believe his point is that a more favorable regulatory regime means more money will be spent in the country of said regime.

But I agree it's likely less of a problem with the EU, New Zealand, Canada etc.

It's actually really simple. First mechanism, there's clinical trials spending. Second mechanism, some official in country X decides to start charging $1M per drug submission, similar to PDUFA. Companies learn that drugs get approved more often/faster in that market and can then be sold in the US and EU. And there you go. Third mechanism, some more sophisticated way to game the system that I can't think of off the top of my head.

I still think some version of reciprocity makes sense, by the way.

First, It's not nationalism, but rather accountability, that makes one choose one's own regulator over a foreign regulator. What are you going to do, fire the foreign regulator for doing a bad job? Conduct oversight hearings on the foreign regulator?

Second, your claim of a billion dollars for drug development assumes there wouldn't be costs of discovery, Phase I, II, and III trials. It also overlooks private cost savings of avoiding errors and costs that are borne by human beings--parents, children, young adults--for drugs that prove to be dangerous because there were no such trials.

Finally, this whole post is a joke and a shill for not confronting the costs of drugs in the US relative to the same drug abroad. If you really understood drug or device regulation, you would also note as a benefit of FDA approval that many countries expedite or automatically approve FDA approved drugs or devices in their country because of the standards we adhere to. So, actually, FDA standards expand the market and facilitate foreign entry.

If a foreign regulator in a developed democracy does a poor job, they will face backlash from their own population and the company that sold the drug would face a worldwide PR disaster. Are there any recent examples of regulators in developed democracies getting things so much more horribly wrong on average than American regulators? Do Canadians routinely cross the border to shop at CVS to escape the perils of shady Canadian pharmaceuticals?

Ricardo, The data from the FDA approved trials is used as data for the foreign approval. On the other hand, a foreign approval may not require the data sufficient for US approval.

That "may" carries a lot of weight in your assumptions.

I don't understand the comment.

"What are you going to do, fire the foreign regulator for doing a bad job?"

Oh, I don't know -- how about suspending the foreign agency from the reciprocity list? Did that option really not occur to you?

So, you admit a drug that would not pass FDA approval, and then take the foreign agency off the list when the drug or device injures people. How about not injuring people first.

"How about not injuring people first."

You ignore the benefits of the process and only focus on the costs. So your view and argument seems skewed to me.

Why do you assume that other regulators are worse at doing their job then the FDA or that FDA approval means a drug will not injure people? It's entirely possible to receive FDA approval but not receive EMA approval and the FDA has approved many drugs that are known to cause harm in clinical trials or that are shown later to cause harm. Also, why hasn't anyone at the FDA ever been fired for approving a harmful drug? The FDA is apparently more interested in firing whistleblowers.

Tell me how I can make the EMA accountable. You seem to believe the EMA is accountable to a US citizen or his elected representative, but the FDA is not. I am sure there are many things the EU is good at, maybe even better at (eg, free education) than we are, but we are just stuck with this stupid idea of accountability and the rule of law. That is not to say the US and EU couldn't agree on common standards, nor is it to say that they couldn't agree on using data from another jurisdiction (which they do.)

What I'm saying is that there is little actual difference in accountability between the two and it does not represent a solid argument for adding billions of dollars of cost to US health care consumers. Your argument that FDA approval is necessary because of an insistence on "the rule of law" is a red herring, there is no constitutional requirement for the FDA to not accept foreign certifications, in fact foreign certifications are accepted for many products that are imported into the US as well as certifications from private organizations. There is also almost no difference in results between the FDA and the EMA approvals. So why exactly do you insist that the FDA reinvent the wheel?

There was a infamous FDA employee who didn't approve any drugs in the 10-20 years he worked there.

Perfect safety record on his watch!

I was simply entertaining your 'what if' hypothetical.

Can you point to any drugs approved by the European or Japanese agencies in recent decades where A) the FDA subsequently rejected the drug AND B) the Europeans and/or Japanese ultimately had to withdraw it from the market for safety reasons AND C) corruption or incompetence on the part of the foreign regulatory agencies was at fault for the initial approval?

Slo, You bear the burden. Show the drugs that were not approved by the FDA but were approved by EMA, and the reasons they were not approved.

You realize that there are benefits as well, correct? Or is that too hard to think about for your populist brain because it's an implicit benefit?

Bill is literally Reddit: the Commenter

First, It’s not nationalism, but rather accountability, that makes one choose one’s own regulator over a foreign regulator. What are you going to do, fire the foreign regulator for doing a bad job? Conduct oversight hearings on the foreign regulator?

How about if we had some agreement that the foreign company could be held liable and sued in American courts for any damages caused by the drug, and the European governments would enforce the judgements?

Finally, this whole post is a joke and a shill for not confronting the costs of drugs in the US relative to the same drug abroad. If you really understood drug or device regulation, you would also note as a benefit of FDA approval that many countries expedite or automatically approve FDA approved drugs or devices in their country because of the standards we adhere to. So, actually, FDA standards expand the market and facilitate foreign entry.

This last part makes no sense. It doesn't help Americans, or lower drug costs in the US, if FDA approved drugs are approved in foreign countries. The way to lower drug costs in the US is to get more drugs approved IN THE US.

"It doesn’t help Americans, or lower drug costs in the US, if FDA approved drugs are approved in foreign countries."

Huh? Of course it does. Drugs priced based on average costs, and average costs are declining in the size of the market.

But even if it didn't affect average costs, Bill's point was that FDA approval helps drug companies. I doubt anyone denies that.

Average costs don't matter for a monopolist (patents) that can engage in price discrimination for each country/jurisdiction.

I think if this were altered from approval to importation we might see a reversal in attitudes even though there would be little difference in effect. Conservatives always want to make drug companies as profitable as possible while progressives want to make drugs as cheap as possible.

Why can't both of those things go hand in hand? Doesn't the company making the product most efficiently usually become the most profitable AND the cheapest?

No. Patents. Efficiency only comes into play once patents have expired and generics proliferate. By that point we're way beyond FDA approval.

Allowing drug imports is probably the best way to get cheaper drugs. But big pharma obviously will oppose that tool and nail.

The other countries will also oppose it. They know the reason they can negotiate better terms is because they agree to not become a shipping route.

What a terrible and useless comment. Especially the part complaining about drug costs. WTF? Also, your view on accountability is simply incoherent. If you think FDA regulators can be trusted because the legislature will 'fire' them if they do a bad job.... good grief.

'It costs well over a billion dollars to get the average new drug approved and much of that cost comes from FDA required clinical trials.'

Laughable indeed - do you honestly think that Sovaldi went through a billion dollars of clinical trials? Well, maybe if Gilead Sciences Inc. were spending a cool 125 million dollars a month - 'The New Drug Application for sofosbuvir was submitted on April 8, 2013, and received the FDA's coveted Breakthrough Therapy Designation, which grants priority review status to drug candidates that may offer major treatment advantages over existing options.

In December 2013, the FDA approved sofosbuvir for the treatment of chronic hepatitis C.'

But what Sovaldi demonstrates is not how the FDA is an impediment, but instead, how drug companies are merely profiteers swimming in a sea of desperate people. 'In Germany negotiations between Gilead and health insurers led to a price of €41,000 for 12 weeks of treatment. This is the same price previously negotiated with the national healthcare system in France, except that additional discounts and rebates apply in France depending on the volume of sales and the number of treatment failures.[54]

The price in the United Kingdom is expected to be GB£35,000 (~US$68,110) for 12 weeks.[55]

In September 2014, Gilead announced that it would permit generic manufacturers to produce sofosbuvir in 91 developing countries containing 54% of the world's HCV-infected population. The company also announced that it would sell a name brand version of the product in India for $300 per course of treatment, approximately double a third party estimate of the minimum achievable cost of manufacture.[56] The leader of one Indian activist group called this move inadequate.[57] Jennifer Cohn of Doctors without Borders and the organization Doctors of the World criticized the price of sofosbuvir as reflecting "corporate greed" and ignoring the needs of patients in developing countries.'

And then there is the question of just how innovative Sovaldi is in any area but pricing - 'In February 2015 it was reported[59] that Doctors of the World had submitted an objection to Gilead's patent[60] at the European Patent Office claiming that the structure of sofosbuvir is based on already known molecules.[61] In particular, Doctors of the World argue that the Protide technology powering sofosbuvir has been previously invented by the McGuigan team at Cardiff University in the UK, and that the Gilead drug is not therefore inventive.'

Let's change just one word in what the Bartley J. Madden Chair in Economics at the Mercatus Center wrote, and see how well it reflects reality - 'We live in a world of pay-offs. Let’s debate the pay-offs. But let’s not engage in magical thinking where there are no pay-offs and “no evidence” that the FDA makes drug development more costly.' And of course the FDA makes drug development more costly - just ask the makers of laetrile how that turned out for them and the desperate, cancer ridden buyers of a product that the FDA had the temerity to block. - and as noted at the link, the FDA stepped in to stop this - 'The promotion of laetrile to treat cancer has been described in the medical literature as a canonical example of quackery,[2][3] and as "the slickest, most sophisticated, and certainly the most remunerative cancer quack promotion in medical history."'

Yes, god damn Gilead for inventing a blockbuster drug that will save the lives of thousands. They are surely bastards. And to think they should make money from saving lives, they're as bad as doctors!

By the way EPO does not allow patents for medical treatments which is why whether the molecule is new matters. In the U.S. it doesn't matter, nor should it.

It's prior_approval. If a German company had created it, he would be lauding its virtues and pointing out how much better German companies are. Since an American company did it, obviously the result is flawed.

One reason I want Medicare to announce "here is what we would pay for a cure for diabetes" is that they would offer a price in the tens of billions.

@JWatts: which is even more amusing because we've now seen proven (by VW) what we already knew, that German companies are corrupt, lying cesspools, as is the entire country of Germany. This is self-evident, and even prior_approval knows it.

Um, no, that was VW, once.

Gilead got a ton of flak for the Sovaldi/Harvoni pricing, but ironically those drugs actually have a quite high QALY per dollar.

Would the QALY go down if the price were only ten times the,labor costs of the drug itself, instead of equal to the labor cost over a lifetime of the only previous treatments?

You are making the argument that government should subsidize innovation to cut costs and put all the profits in the subsidized government picked winners.

Why not argue that at the first suggestion that the drug sofosbuvir might be a cure result is the drug flooding the market at a dollar a pill to be tried by people with hep C to see if it's a cure for them?

I have no idea what this means.

No one knows what it means. Who among us can fathom the wisdom of the mulp? The mulp expounds, he does not explain. Any failure to comprehend the mulp is your fault, not his. Or rather, you simply do not have the understanding necessary to follow the deep wisdom. And thank whatever god you pray to that this is so, for the mulp has access to knowledge so vast and terrifying that we mere mortals would blanche with fear to know it.

If it were a dollar a pill, more people could access it and more QALY years would be gained.

But many people provide evidence that Gilead did not invent the drug.

Yes, Gilead paid for the clinical trials that proved the drug invented by others is a 99% cure today for one disease, but why does that justify a $50,000 monopoly profit on a hundred dollars of labor costs, plus perhaps a thousand in clinical trial labor costs per treatment the first few years?

Is the $50,000 each treatment going purely to funding labor costs of more clinical trials of chemicals they bought?

It's not a secret. Gilead bought Pharmasset in 2010 if I recall correctly, and that's how they got the sofosbuvir molecule. So what?

It tells you a lot that a Harvard professor of medicine would blatantly lie like that, when interviewed for publication.

The system you are trying to improve has very little to do with helping people. It has much more to do with generating large incomes for bureaucrats. As such, good luck with your supposed "improvements," because the ones in charge don't quite see them that way.

I was impressed by Aaron Kesselheim's credentials. Let's give the Devil his due: does Kesselheim know something AlexT does not? Perhaps Dr. K is not lying? Perhaps the cost of new drugs is in lengthy trials that would occur even if there was no FDA, because of the possibility of medical malpractice suits? Food for thought.


Aaron Kesselheim
Research Associate, Harvard University
Department of Health Policy and Management

Dr. Kesselheim’s research focuses on the effects of intellectual property laws and federal and state policies on prescription drug and medical device development, the regulatory approval process, and the costs, availability, and use of therapeutics both domestically and in resource-poor settings. He has also investigated how other issues at the intersection of law and public health can affect health care delivery, including health care fraud, expert testimony in malpractice cases, and insurance reimbursement practices.

Another possibility is that Vox mangled the interview.

There are a number of issues all entangled here; but if you think that the regulators are interchangeable, that's a bold claim on many bases. If they aren't interchangeable, then there may be legitimate reasons for different outcomes to the same application.

That's where the real savings are, then - in allowing a 'common application' with regard to data format, etc. It can be tricky, because of languages and regulatory details. Also, it won't effectively enable 'selective optimization.' Further, if one of the reasons that regulations differ is because treated populations differ demographically, then a drug might actually be safe in one country and not in another - and so the clinical trials will need to be run twice (or n times) anyway.

None of which is to suggest that the FDA is doing the best possible job; but to say that the fixes are rarely as easy as people suggest.

The FDA is doing the best job the drug company payments for FDA employees can buy, which is the deal that conservatives pushed to cut tax funded spending and putting all costs for government approval that leads to high profits on those getting the profits.

This is the change made to patent approvals to cut cost to taxpayers of the patent system - if you want a patent you must pay the employee to tests whether your patent application meets the Constitutional requirements.

I agree with point that reciprocity will be beneficial, but i am not sure about the " 140,000 to 310,000" LY saved calculation. Looked a bit vague to me. Also letting Congress have the power to overrule FDA decisions seems dangerous (According to Alex, there only benefits, nor harms loool) . Also if this claim in the Vox article is true ( "Last time we looked, the agency had approved 96 percent of the applications that came its way in 2015.") I kind of wonder if in practice there will be much benefit for patients.

With numbers like that, its important to note that they don't say how quickly drugs are approved (for example, some folks may need to re-apply or something like that) or include drugs that may have marginal benefits but aren't worth the company's hassle to try the application process.

If the approvals process is well-understood bureaucracy, you'd expect a large approval rate. If everyone knows that to get FDA approval, you have to have so many studies with such-and-such parameters, and demonstrate X results, you shouldn't be surprised when all the applications have that.

If the approvals process is checking off boxes, rather than making scientific judgements about whether treatment X is really worthwhile, you expect a very high approval rate.

If you change the bureaucracy to make it easier to get approval, and the change is similarly well-known, you'll still get 86% (or whatever) approval rates with the easier bureaucracy, but there will be fewer studies and a smaller application packet.

Tabarrok might be interested in this example of reciprocity: Physicians in the U.S. are regulated and licensed by the states. That has worked well when physician services were delivered locally, but it doesn't work so well now that telemedicine is changing the way physician services can be delivered (i.e., from a distance). The problem arises when the patient is located State A and the physician is located in State B and the physician isn't licensed in State A. What's the solution? The physician could seek a license in every state in which he delivers physician services via telemedicine, but that's not a practical solution. The model telemedicine act provides the solution: reciprocity. If State A and State B agree to reciprocity, then physicians in both State A and State B can provide physicians services to patients located in both states. So what's the problem? The physicians in State A may not wish to grant reciprocity to physicians in State B because, for example, they fear that the physicians in State B would end up with all the patients in State A. Thus, a promising technological advancement (telemedicine) can't be implemented across state lines because physicians want to protect their turf. So what about reciprocity in the approval of pharmaceuticals? Who, exactly, opposes reciprocity: the regulators in the different countries or the pharmaceutical companies in the different countries? Tabarrok would blame the regulators, which is not surprising. What about the pharmaceutical companies who want to protect their turf? Indeed, would reciprocity result in a race to the bottom, the country with the least amount of regulation the place where all approvals would be sought? Would pharmaceutical companies relocate in that country to gain an advantage in the approval process?

"Indeed, would reciprocity result in a race to the bottom, the country with the least amount of regulation the place where all approvals would be sought?"

No. The number of foreign agencies whose approvals might be accepted reciprocally is small. For starters, the list might include just the EMA (Europe) and PMDA (Japan).

Why exclude Russian government approval? Indian government approval?

Russians have an instant, permanent, cure for alcoholism, why not approve it in the US? Those Russians who deliver the cure say it's 100% successful.

I look forward to your evidence of Pharma companies who oppose reciprocity.

Who can afford the approval process? It's a barrier to entry that gives the existing companies an enormous advantage, and helps explain why pharmaceutical companies devote most of their resources to making small changes to existing drugs rather than developing new ones (because they don't have to). With reciprocity the pharmaceutical companies might lose their advantage. Rubio and Lee introduced the legislation with a wink and a nod to the pharmaceutical industry. As my long-time assistant would say: I may have been born at night, but I wasn't born last night.

From Dr. Kesselheim's bio that the Vox piece has a broken link to ( "Dr. Kesselheim’s work is also currently funded by the FDA." You'd think something like that should be disclosed in the Vox piece.

Full disclosure: Vox is garbage.


But let's give Alex props: he clearly labeled the article as "Vox" so the reader could eschew the link if he/she wanted. This is a great practice and should be repeated.

The best practice would be to *never* link to Vox, but this is good enough.

James Taranto refers to Vox as a "young adult website.

I emailed both the author and an editor, let's see if they add a disclosure!

(Who am I kidding, the chances are slim to none - Vox is a rag)

I think you sketch a world where drugs are good, and approval only slows them down.

It actually seems opaque which drugs on and off the market are good. Yesterday's (speculative) report that mom's antidepressants increase autism risk is a case in point.

We probably don't live in a world where more approval means more health ..but we definitely live in a world where it is hard to tell.

Possibly the best 'big picture' comment yet made.

I used to work in portfolio management at a major pharma company. We helped management make decisions about what got funded in the pipeline. We had an internal hurdle rate that was the cut-off in terms of acceptable returns on R&D investment.

Because patent expiration is a fixed date, every month shaved off approval process is an extra month of having the drug on the market AT PEAK SALES (drugs have an uptake curve on the front end that takes 3 to 5 years; moving that forward means more time between reaching the peak sales plateau and patent expiry). Even with high discount rates, that can translate into significant value, and can be the difference between a drug being worth the risk or not (meeting the internal hurdle rate). Potential blockbusters (even the ones lamented as "marginal improvements" over existing drugs) will always get funded. The question is whether the smaller market drugs are worth the expense and risk. Anything that boosts profitability will expand the universe of what's worth pursuing.

So, if we give reciprocity, you are arguing we should eliminate the patent extension that was designed to compensate for FDA approval.

So, drug company investments are driven totally by monopoly profits, not on investment in labor and production labor to get a return of 8% of actual equity on drug production?

This means the only source of 95% of drugs is foreign drug manufacturers who make money by earning a return on equity over total labor costs to make the drugs, the failed economic classical free market capitalism system?

Why yes. You described it so perfectly that you need not comment on this web site any more.

There is no evidence or reason to think that the cost of clinical trials is "much" of the putative "well over a billion dollars" that it costs to develop a drug. Nor did the FDA invent its requirements out of thin air. Instead, Congress has demanded that drugs be shown by rigorous evidence to be efficacious. Perhaps Alex needs to be reminded of the value of randomization in causal inference, or of the role of samole size in determining statistical power, but it is very odd to act as if such evidentiary requirements are just stuffy bureaucratic red tape. Finally, Alex's posts on the FDA always seem to be completely uninformed by the actual evidence on how the FDA uses various accelerated approval mechanisms where the evidentiary standards (unfortunately, some might say) are already as low or lower than what Alex recommends, apparently under the impression that he is proposing a reform that isn't already occurring.

I think you're missing Kesselheim's point, badly. His quote in context is explicitly comparing the FDA to other "regulatory agenc[ies]... in the world". So he is arguing that the FDA does not impose extra burdens to innovation compared to the baseline of elsewhere-existing regulatory regimes. Maybe he's wrong, but he's not engaging in magical thinking that there are no tradeoffs involved in regulation.

Even so, this implies massive duplication of effort.

Why is it so easy to imagine that Senior Fellow Bartley J. Madden (Heartland Inst.) is fully aware of this? In part due to his apparent ongoing devotion to this subject for almost a decade - 'Since his retirement as a Managing Director of Credit Suisse in 2003, Madden has written articles that developed a market-based system for access to not-yet-FDA-approved drugs. In May 2007 Heartland published a booklet, More Choices, Better Health: Free to Choose Experimental Drugs, that summarizes this work.' From a link at concerning Bartley-Madden/28403219

A booklet that can be downloaded as a PDF from google -

And as noted by Madden himself in 2012, this booklet received the following treatment - 'The Heartland Inst. plans a major campaign to promote the ideas described in my booklet "More Choices, Better Health - Free to Choose Experimental Drugs" The first steps will include:

* 24,000 booklets to be mailed in the U.S., including all elected state and federal officials and those in the news media who focus on health care

* 10,000 booklets to be mailed to influential leaders worldwide

* electronic files of translations of the booklet will be made available'

Ideas, with the proper funding, never seem to go away, regardless of their merit. Comments on that reality, however....

You're the expert on things that never seem to go away.

Philipson et al. (2008) implicitly assumes that patients WTP is closely related to a drugs (health?) benefit. Patients and any individual pharmacist, med. doc., etc. are unable to estimate the efficiacy of any drug, which is why they rely on the FDA. Despite the FDA's stringent tests, over subscribing and overtreatment are important health issues. Even an effiective drug that creates significant health beneftis to correctly medicated patients might still produce a negative consumer surplus effect if it is over suscribed (e.g. opioids, antibiotics, cancer screenings). As a higher sales volume at the same price increases the producer surplus, the pharmaceutical industry profits from excessive subscribtions and treatment. Tabarrok and Philipson et al. should be more sceptical in applying the concept of welfare surpluses to medicine in general.

Suggesting that it costs over a billion dollars to get the average new drug approved fundamentally and woefully misunderstands the nature of drug development. It gives the impression of a vast range of tests and trials on a single drug. But it's not like that at all. The clue is there in the linked article:

“Because the R&D process is marked by substantial technical risks, with expenditures incurred for many development projects that fail to result in a marketed product, our estimate links the costs of unsuccessful projects to those that are successful in obtaining marketing approval from regulatory authorities.”

So the vast majority of the cost isn't the trial and approval process for a single drug that works. It's the development and trial costs for dozens of other drugs that fail. Here's a massively simplified version.

Suppose you have a three stage development process. You develop and try a hundred new drugs at the first stage. This costs $4m for each. Total cost so far: $400m. Most of these will either do nothing, be no better than existing drugs, or be actively dangerous.

Ten of the drugs show enough potential and are safe enough that they're worth carrying on with. Second stage development and trials on these costs $30m each. Total cost so far: $700m. From these ten, only two actually seem to do enough to be worth continuing into the third stage.

Third stage trials cost $150m each, and one of the drugs turns out not to replicate, in larger trials, the benefits it seemed to show at stage 2. Only one is approved. Total cost to get that one drug approved: $1billion.

But most of this cost isn't down to some set of mythical uniquely costly FDA (or EMA) tests. It's down to the fact that drug development is hard, and most drugs fail - not in the sense of failing to get approval but simply in the sense of failing to do what you hope they will. Of course there's room for improvement in regulation, and reciprocity may be part of the answer, but using loaded language like "a billion dollars to get the average new drug approved" really won't help.

Thank you. That makes a lot of sense. It doesn't seem remotely possible that developing a new drug costs $1billion.

He makes an extremely good point, but I don't see why it should be surprising to find that a new drug would cost a lot of money to find when the low hanging fruit have presumably been found.

Evidence that Alex should consider:

"There is no evidence or reason to think that the cost of clinical trials is “much” of the putative “well over a billion dollars” that it costs to develop a drug. " That seemed way off, so I looked it up.

Around $30 to $40 million for phases I, II and III and about the same or post approval phase IV. So maybe $60 to to $80 million.

Baked into the higher number are all of the drugs that fail their trials. The number isn't a billion either it is more like 1 billion.

More like 2 billion.

So, the requirement to test drugs is a huge burden preventing cures from reaching the market because of the high costs of testing the 9 out of 10 drugs that do not work?

If 10 out of ten drugs get put on the market, all ten of these drugs will be cures justifying the $10,000 price, or after patients and taxpayers pay $10,000 for a drug that does not work, no one will buy it anymore and that means the drugs that work will be found without any cost to the drug company shareholders, because patients and taxpayers will have paid the costs in cash and lives?

It's not just finding out that it works. Even in cases where we are just about sure it will work -- generic medications that have the same molecule -- there are still bioequivalence tests that keep extra generic producers of small-time things like pyrimethamine out of the market.

So we are talking about up to 8% of spending. That's not nothing, but it is quite misleading to describe 8% as "much" of the cost.

For giant drugs with huge markets, it's probably not a huge problem. For smaller drugs with smaller markets, like pyrimethamine, it's a bigger problem.

And then you'll get people posting about how all the pharma companies only want to make boner pills.

FDA requires drugs to be "safe and effective," yet the meaning of that phrase is modulated by recognition of risk-vs-benefit calculations. Arguments for less regulation invariably seem to lead to the terminally-ill patient who wants to try something that's not been approved by FDA, and who could reasonably argue that such patients should not be able to do so?

BUT the real issue is, "who pays and how much"? In the USA government pays for an ever-increasing percentage of medical treatment (including drugs) and is usually obligated to pay for any drug treatment FDA has approved as "safe and effective." Which gives government very little negotiating power when buying drugs. Government doesn't pay for apricot-pit nostrums, but it does pay for many other cancer drugs of dubious effectiveness. Drugs can often get FDA-approved if they show improvement in diagnostics (such as tumor size reduction) even if they can't show direct measures of patient benefit, such as increased survival time or improved quality of life.

Few argue that such drugs should be denied to patients with malignancies that are not responding to other treatments, yet these drugs are often insanely expensive and may do little for those who receive them. Which perhaps raises the question of whether taxpayers should be obligated to pay for them, or whether government should be able to make take-it-or-leave-it offers for these drugs, considering the weak case that they actually offer significant value.

Which is to say, regulatory reciprocity (especially if only for a trial period, and not to be renewed without adequate assessment) might be a useful tool both to increase the range of drugs available legally in the USA and to reduce costs. BUT there are many reasons beyond high regulatory hurdles for high drug prices in the USA, and these, too, could/should be attacked.

Which is politically difficult, as AARP is a mighty lobby that is rightfully feared by politicians, and "I should have a right to use drugs of questionable effectiveness, especially if there are no alternatives" inevitably segues into "AND someone must pay whatever it costs to obtain them for me." Which leaves the "someone" with little or no power to negotiate price.

Very good comment.

very little negotiating power when buying drugs.

Medicare has literally no negotiating power because it was explicitly forbidden by the law that created the prescription drug benefit.

I believe the VA does, however. So the solution is for everyone to join the military.

That was a bizarrely bad Vox article. "Time doesn't increase costs". Assuming for a billion dollar clinical trial and a (lowball estimate of ) the annual cost of capital of 8%, and a six year approval period. The time value of money works out to $600 million.

For folks who aren't in finance, to arrive at how much money a pharmaceutical company must expect a drug to make in order to proceed with research you take the costs of the R&D etc, then add that $600 million. This is actually an underestimate, since I've ignoring the time it takes to do the R&D and for the revenues to come in. In other words, you could charge ~40% less for a drug approved in a day than for a drug approved six years later. Or target a 40% less common disease, etc.

When did the Fed hike rates to 6%?

I remember 8% being the standard return on investment for utilities regulated by the PUCs, utility bonds paid 6-8%, bank savings paid the 4.5% Regulation Q max, and US Treasuries were 5%.

So, now with cash sloshing around looking for investments, interest under 2% everywhere, yet return on investment is still 8%??

Why shouldn't return on investment be more like 4%? Why are people with cash willing to earn zero on the money instead of say 4% but hold out for 8-10-15-20% returns, constantly blaming government for failing to subsidize them or let them take high profits by putting all the risks on government, customers, society, the world?

Even when conservatives write the rules for government, conservatives blame the government rules they wrote.

As a liberal, I would have the government spend a trillion on drug research based on the high cost patients and problems, test the drugs to find the drugs that cut the cost of treatment and cure, and then contract with private drug makers to sell these government drugs at the labor cost of capital and operations plus 8% ROE. There would be zero economic profit to the private sector, only to the public in the lower cost of health care.

"So, now with cash sloshing around looking for investments, interest under 2% everywhere, yet return on investment is still 8%?? "

So mulp, how much money will you loan me at 2% for a 6 year period?

I got the 8% from this NYU website: Honestly, the 8% is probably a substantial underestimate for our purposes as it's the overall pharma cost of capital rather than the drug development specific discount rate. The 8% includes many less risky activities, like investing in marketing.

To be fair, you are correct that there would be ENORMOUS savings to society if we could swap the federal cost of capital (~2% as you noted) for the private cost of capital. This is actually one of the major (and underappreciated) benefits of funding drug development via prizes paid upon FDA approval. Such as system retains most of the advantages of the market with respect to identifying promising drugs, but functionally swaps the cost of capital down to the federal rate during the 10 to 20 years of patent life left on a drug upon approval.

Perhaps it would be good for targeting a 40% less common disease, but I doubt that very many people at all believe that pharmaceuticals would pass on the savings if they could get things approved faster.

Costs of regulation should be calculated net from the benefit. If you think the market will properly account for risk in new drugs, well, that seems magical to me. I would be interested in seeing an example. If you cant think of an example where lack of regulation led to bad outcomes you are not being honest.

There is no evidence showing negative costs net of benefits. In order to explain things so concisely vox skips a detail or two or three. This doesn't amount to magical thinking from my point of view.

My mother worked in statistical analysis of clinical trials for a drug company. She told me that, since the FDA has the most exacting standards, they ran their trials to those standards. Partly because of this, I get the impression that they run one set of trials (per stage, of course) and submit those results to all the agencies. If this is true, it doesn't square with the author implying that there's much duplication of effort - maybe he's referring to the bureaucratic overhead of submitting the trials rather than running them?

"much of that cost comes from FDA required clinical trials."

Is this Alex calling for his doctor to work with drug companies and charging Alex to be a guinea pig for really great drugs that come straight from a drug lab?

Alex seems to think Walter White is the ideal drug maker, free of all government regulation and making truck loads of profit.

"Trust me with your life, I'm a great drug maker totally focused on making money, so you know the high price I charge is worth it to me."

Walter White did pride himself on making a high quality product.

Vox interviewed a Harvard Professor and they don't like an R-submitted bill in Congress, shocker!

This is exactly the sort of issue on which libertarians have a lot to add to public debate.

While I don't agree completely with the post, and find much of the Cruz-Lee proposal dubious and arbitrary, there is much of value in Alex's arguments. talked a lot about a "new kind of journalism". I always mistrusted them and thought that a lot of articles sounded like they are written by pretty ignorant left-wing partisan hacks. And they are. This article is just one sad example. Thanks to Alex Tabarrok for the HT.

VOX is like any other middle brow news source, the irritating part is the pretence that they are impartial.

More actual evidence:

Alex, I would really like to hear your responses to Derek's concerns, as I think they are far more substantive than the Vox article. Although a supporter of the idea in general, I find his concern about a race to the bottom very sensible. And the congressional 'veto' is, of course, nuts.

One depressing feature Tabarrok's posts is that once you scratch the surface of his drug regulation arguments, it becomes obvious that he is arguing in bad faith.

For example, the Philipson et al. study says NOTHING about dynamic effects of regulation on innovation and bringing new drugs to market. The authors say as much in the discussion section. Yet, Tabarrok uses the paper to argue this exact point....

If you are going to accuse someone of arguing in bad faith it behooves you to read the post carefully. I quoted the Philipson study on speedier approvals under PDUFA and lives saved. I did not use it to argue for dynamic effects. But immediately following the discussion of Philipson, I discussed the Vernon et al. study on dynamic effects, going so far as to highlight that the dynamic effects are "over and above the benefits of more rapid consumer access." Thus, I discussed static and dynamic effects, one study for each.

It seems a bit peculiar that our progressive brethren, who're so very eager to have the US adopt the health-care systems of the European nations, are so vehemently opposed to our accepting their decisions on drug approvals.

Got a mirror?

Judges making decisions based on European law = an evil socialist theft of U.S. sovereignty and submission to the U.N.
An executive agency making decisions based on European law = common sense.

This is a bait and switch. PDUFA just gave FDA more money to review applications faster, it did not lower the evidentiary standard, which is what Alex wants to do.

Hoisted from comments from a post on this topic several days back:

steve December 14, 2015 at 8:19 pm
Written by Grace Rattue Published: Fri 18 May 2012 at 2pm PST email8 Although the drug approval process in the U.S. has been perceived as too slow, the nation approves new drug treatments faster and earlier than Europe and Canada, according to researchers at Yale School of Medicine.

The study, conducted by Nicholas Downing, a second-year medical student and Joseph S. Ross, M.D., assistant professor of internal medicine at Yale School of Medicine, found that the median total time to review was: 322 days at The U.S. Food and Drug Administration (FDA) 366 days at The European Medicine Agency (EMA) 393 days at Health Canada The teams findings was published online by the New England Journal of Medicine. Regulatory review is the final step in the process of bringing new drug treatments from the lab to patients. As a result, it is vital that the review process is efficient in order to get promising new therapies to patients sooner, while ensuring drug safety.

Downing explained:

“The perception that the FDA is too slow implies that sick patients are waiting unnecessarily for regulators to complete their review of new drug applications.”

The researchers reviewed drug approval decisions of the FDA, EMA, and Health Canada between 2001 and 2010 in order to determine the time it took to review novel therapeutics and timing of key regulatory events. The team selected Europe and Canada as a comparison due to the fact that they face similar pressures to approve new therapies quickly while ensuring drug safety.

Downing said:

“Among the subsample of drugs approved for all three regulators, the FDA’s reviews were over three months faster than those of the EMA or Health Canada. The total review time at the FDA was faster than EMA, despite the FDA’s far higher proportion of applications requiring multiple regulatory reviews.”

Downing continued:

“Examining novel drugs approved in multiple markets, we found that 64% of medicines approved in both the U.S. and in Europe were approved for U.S. patients first, and 86% of medicines approved in both the U.S. and Canada were also approved first in the U.S.”

- See more at:

While Kesselheim is clearly exaggerating his case, Alex is the one who is engaged in more "magical" thinking---to believe that drug development is so expensive just because of the regulation and that the "market" will just fix anything.

Alex appears not to appreciate the technical issues that challenge drug development and testing.

Drug development is getting more expensive because the "low hanging fruit" has been picked. Many of the drugs are either replacements for ones that are successful or address conditions where demonstrating effectiveness is difficult.

In many of these cases it is very difficult to demonstrate whether a drug works and in many cases the drug only works in less then half of the cases and "works" is add a few months to someone's life or make them feel a little better. Also, many of the drugs may have to be taken for years, so it is a challenge to know whether the long term harm exceeds the benefits.

And seriously to all liberatians: How the hell can I find out whether some drug that my doctor prescribes is more likely to help me then hurt me.

Sorry, modern life has its benefits, but it is complex.

Hey Alex, responded to your comments about my piece in Vox. As I explain in the post, I don't think speeding up or whitting back on the regulatory process is going to solve the innovation problem you describe.

Can someone here name a single drug that was denied by the FDA or EMA yet produced excellent results in other countries?

Controlling for other factors such as pharmaceutical profitability and cash flows, we estimate that a 10% decrease (increase) in FDA approval times leads to an increase (decrease) in R&D spending from between 1.4% and 2.0% - See more at:

Counter argument here. Suppose the industry bet big on some hot 'killer idea' (say targeted cancer drugs or immunotherapies). The first set of drugs in this class do wonders, so great clinical trials are stopped early because the results are so good. FDA approves fast, profits are huge, R&D increases dramatically. As time goes on, the idea starts getting exhausted. New drugs in the class only seem to have marginal benefits, sometimes the statistics justifying them as an improvement over existing drugs seems stretched. FDA approvals start taking longer, profits seem to be smaller since the original wonder drugs are now going generic and insurers are skeptical about paying premium prices for the newer batch of brand name drugs that seem to be only slightly better. As a result R&D starts to dry up.

There are certain cases where it makes a ton of sense. Take cholera and tick borne encephalitis. Both are common diseases elsewhere, and both have perfectly functioning vaccines, not available in the US legally because the diseases are rare here. (They are available in Western Europe.) They are on WHO's list of essential medicines.

They will possibly never be approved in the US simply because the diseases are so rare here. But what are travelers supposed to do? Surely it would make sense to get travel immunizations before going to a foreign country. In the case of yellow fever, there are special FDA clinics approved to give the vaccine, but for cholera, people have to travel to Canada or elsewhere to get the vaccine if they're traveling to a country where it's endemic. It's ridiculous.

Comments for this post are closed